[GIT 00/19 v2] Renesas ARM based SoC pinmux updates for v3.12

[Olof Johansson]

On Sun, Aug 18, 2013 at 01:55:57PM +1000, Simon Horman wrote:
> On Wed, Aug 14, 2013 at 12:23:28AM -0700, Olof Johansson wrote:
> > On Wed, Jul 24, 2013 at 09:32:14AM +0900, Simon Horman wrote:
> > > Hi Olof, Hi Arnd,
> > > 
> > > please consider the following pinmux updates for v3.12.
> > > 
> > > Difference with v1 of this pull-request:
> > > * Remove bogus i2c portions of the first patch of the series.
> > > * Correct addressing mode used in dts of the first patch of the series.
> > > * Removal of egg from face: wip
> > > 
> > > The following changes since commit ad81f0545ef01ea651886dddac4bef6cec930092:
> > > 
> > >   Linux 3.11-rc1 (2013-07-14 15:18:27 -0700)
> > > 
> > > are available in the git repository at:
> > > 
> > >   git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas.git tags/renesas-pinmux-for-v3.12
> > 
> > Pulled, but I noticed this. Can you please follow up with bindings revisit in
> > time for 3.12 final (i.e. if we need to take it after merge window that's
> > probably OK):
> > 
> >         pfc: pfc at e6060000 {                                                     
> >                 compatible = "renesas,pfc-r8a7790";                             
> >                 reg = <0 0xe6060000 0 0x250>;                                   
> >                 #gpio-range-cells = <3>;                                        
> >         };                                                                      
> >                                                                                                                
> >         sdhi0: sdhi at ee100000 {                                                  
> >                 compatible = "renesas,r8a7790-sdhi"; 
> > 
> > 
> > I.e. different ordering on the compatible values looks a bit uncoordinated.
> 
> Can I confirm that you are referring to the fact that
> assume you mean that "-sdhi" follows "r8a7790" whereas "pfc-" precedes
> "r8a7790"?
> 
> If so, I'll work with Laurent and Guennadi (both CCed, both on vacation),
> and Magnus to resolve this inconsistency.

Yes, exactly. Some platforms use <vendor>,<soc>-<ip>, others
<vendor>,<ip>-<soc>. I don't care which way is used, but it would be good if
everybody used the same convention to make it easier for developers.

Preference for direction is something I leave up to the device tree maintainers
to declare. ;-)


-Olof